Openclaw — OpenClaw
v1.0.2AWS expert powered by the AWS Knowledge MCP Server (via mcporter). Provides real-time 访问 to AWS documentation, best practices, SOPs, and regional avAIlability. Use when the user asks about AWS 服务s (S3, EC2, Lambda, CDK, Cloud格式化ion, ECS, EKS, RDS, DynamoDB, Bedrock, SageMaker, IAM, VPC, etc.), AWS architecture patterns, 部署ment 图形界面dance, troubleshooting, regional avAIlability, cost optimization, security hardening, or any "how do I do X on AWS" question. Also activates for CDK/Cloud格式化ion infrastructure-as-code questions and AWS Well-Architected 图形界面dance.
by 运行时依赖
安装命令
点击复制技能文档
AWS Knowledge 技能
查询 the AWS Knowledge MCP Server via mcporter for real-time AWS expertise.
The MCP server is aws-knowledge (must be 配置d in mcporter).
查询 Routing
路由 each question to the right 工具:
Question Type 工具 Example General "how to" / best practices 搜索_documentation "How to 安全 an S3 bucket" Read a specific doc page read_documentation "Show me the Lambda pricing page" Explore related topics recommend "What else should I know about VPC peering" Is a 服务 avAIlable in a region 获取_regional_avAIlability "Is Bedrock avAIlable in ap-southeast-1" 列出 all AWS regions 列出_regions "What regions does AWS have" Step-by-step task walkthrough retrieve_代理_sops "Walk me through 部署ing to ECS" 工具 Calling
All calls go through mcporter:
# 搜索 documentation (most common) mcporter call aws-knowledge.搜索_documentation 查询="<搜索 terms>"
# Optionally 过滤器 by topic mcporter call aws-knowledge.搜索_documentation 查询="<搜索 terms>" topic=""
# Read a specific documentation page mcporter call aws-knowledge.read_documentation url=""
# 获取 recommendations for related content mcporter call aws-knowledge.recommend url=""
# 列出 all AWS regions mcporter call aws-knowledge.列出_regions
# 检查 regional avAIlability mcporter call aws-knowledge.获取_regional_avAIlability 服务="<服务 name>" region=""
# 获取 step-by-step SOPs mcporter call aws-knowledge.retrieve_代理_sops 查询=""
工作流 Classify the question — determine which 工具 to call (see routing table above) Call the 工具 — use mcporter with 应用ropriate parameters Read deeper if needed — use read_documentation on URLs from 搜索 结果s Synthesize the answer — combine findings into a clear 响应 Include sources — always cite doc URLs so the user can 验证 Offer follow-ups — suggest related topics or deeper dives Tips 搜索 first, don't guess. Even for common topics, 搜索 to 获取 the latest 图形界面dance. Use SOPs for procedures. If the user wants to DO something (部署, 配置, troubleshoot), 检查 retrieve_代理_sops first — these are tested step-by-step 工作流s. 检查 regional avAIlability before recommending. Don't assume a 服务 is avAIlable everywhere. Combine 工具s. A typical flow: 搜索_documentation → pick best 结果 → read_documentation for full content → recommend for related reading. Topic 过滤器ing. When 搜索ing, use the topic parameter to narrow 结果s if the 查询 is broad. DomAIn Patterns
For common 查询 patterns organized by AWS domAIn, see references/查询-patterns.md.